Here, we present a simple technique to assess environmental antimicrobial resistance (AMR) by enhancing the proportion of low-molecular-weight extracellular DNA. Prior treatment with 20%-30% PEG and 1.2 M NaCl allows detection of both genomic and horizontally transferred AMR genes. The protocol lends itself to a kit-free process with additional optimization.
Environmental surveillance is recognized as an important tool for assessing public health in the post-pandemic era. Water, in particular wastewater, has emerged as the source of choice to sample pathogen burdens in the environment. Wastewater from open drains and community water treatment plants is a reservoir of both pathogens and antimicrobial resistance (AMR) genes, and frequently comes in contact with humans. While there are many methods of tracking AMR from water, isolating good-quality DNA at high yields from heterogeneous samples remains a challenge. To compensate, sample volumes often need to be high, creating practical constraints. Additionally, environmental DNA is frequently fragmented, and the sources of AMR (plasmids, phages, linear DNA) consist of low-molecular-weight DNA. Yet, few extraction processes have focused on methods for high-yield extraction of linear and low-molecular-weight DNA. Here, a simple method for high-yield linear DNA extraction from small volumes of wastewater using the precipitation properties of polyethylene glycol (PEG) is reported. This study makes a case for increasing overall DNA yields from water samples collected for metagenomic analyses by enriching the proportion of linear DNA. In addition, enhancing low-molecular-weight DNA overcomes the current problem of under-sampling environmental AMR due to a focus on high-molecular-weight and intracellular DNA. This method is expected to be particularly useful when extracellular DNA exists but at low concentrations, such as with effluents from treatment plants. It should also enhance the environmental sampling of AMR gene fragments that spread through horizontal gene transfer.
SARS-CoV-2 and its aftermath underlined the importance of environmental surveillance in monitoring and predicting infectious disease outbreaks1,2. While viral pandemics are apparent, the rise of antimicrobial resistance (AMR) is often described as an insidious pandemic and one that constitutes a leading public health concern across the world3,4. Consequently, there is an urgent need for coordinated strategies to understand the evolution and spread of AMR. Water bodies, as well as wastewater, can serve as reservoirs for both pathogens and AMR5,6,7,8. Shared water sources are, therefore, a potent source of disease transmission among humans, particularly in low and middle-income countries (LMIC) where poor hygiene and over-population go hand in hand9,10,11. Testing of water sources has long been employed to assess community health12,13,14. Recently, wastewater from urban sewage treatment plants proved a good advance indicator of COVID cases in the clinic1,2,15,16,17,18.
Compared with monitoring specific diseases, detecting and tracking AMR in the environment poses a more complex problem. The large number of antibiotics in use, diverse resistance genes, different local selection pressures, and horizontal gene transfer among bacteria make it difficult to assess true AMR burden and, once assessed, to correlate it with clinical observations19,20,21,22. As a result, while concerted surveillance of clinical AMR is being carried out by several organizations across the world3,23,24, environmental AMR monitoring is still in its infancy, reviewed in19,25,26.
In recent years, different methods for tracking environmental AMR have been reported5,27, reviewed in28,29. The starting point of most of these is the extraction of good quality DNA from heterogenous environmental samples, in itself a challenge. Additionally, environmental DNA is typically fragmented because of exposure to hostile surroundings. Fragmented extracellular DNA has long been recognized as an important reservoir of AMR genes (reviewed in30,31,32), with the added potential to enter and leave bacteria via horizontal gene transfer. Hence, it is important that any protocol that aims to measure AMR burden in the environment should sample linear and low-molecular-weight DNA as best as possible. Surprisingly, there has been little focus on developing methods specific to high-yield extraction of linear and low-molecular-weight DNA: this work focuses on addressing the gap.
A common and simple method to precipitate DNA is to combine polyethylene glycol (PEG) and salts such as sodium chloride (NaCl)33. PEG is a macromolecular crowding agent used to achieve size-specific precipitation of DNA fragments34,35. The lower the PEG concentration, the higher the molecular weight of DNA that can be efficiently precipitated. Many studies have used PEG during environmental extraction of DNA and RNA1,2 (summarized in Table 117,33,36,37,38,39) either in the final step 33,36,37or to concentrate large water samples for extraction of viral particles as with SARS-CoV-215,40. In the current work, it is found that the PEG concentrations used previously for environmental DNA extractions (largely determined by viral surveillance protocols) do not capture low-molecular weight linear DNA. Therefore, they lose out on sampling short DNA fragments and are unsuitable for assessing AMR content accurately. This study has exploited the properties of polyethylene glycol and sodium chloride to effectively precipitate low-molecular weight linear DNA fragments at a high yield that can, in the future, lead to a cost-effective DNA extraction method. This method can be used to enrich the proportion of fragmented and low-molecular-weight DNA from complex natural samples, thus capturing a more accurate picture of environmental AMR. With a little further refinement, the technique lends itself to easy and low-cost application by local municipal corporations and other government bodies to use as a surveillance tool with minimal technical training.
1. Wastewater sampling
2. DNA extraction from wastewater samples
3. Precipitation of polymerase chain reaction (PCR)-amplified linear DNA to check DNA recovery across a range of molecular weights
Establishment of a protocol for high-yield extraction of DNA from wastewater samples
A modified version of previously established protocols was used for the extraction of high-quality DNA and RNA from water samples17. The samples were sourced from open drains as well as sewage treatment plants in the Delhi-NCR region of North India. After pre-processing using PEG and NaCl (Figure 1), the samples were processed through kits for extraction of DNA from soil, and water. In all cases, a prominent high-molecular-weight band likely corresponding to intracellular bacterial DNA and a background smear, as is typical of environmental samples37, was observed (Figure 2).
Lack of efficient DNA extraction from sewage treatment plant (STP) effluent
Although a reasonable yield of total DNA was obtained from open drains and STP influents (Table 3), no DNA could be obtained from the final treated effluent; problems with low yield are described in previous work as well41 (Figure 3). Post-treatment, which includes chlorination of wastewater and, in some cases, both UV and ultrafiltration treatments42, the microbial load is expected to be low, and any residual DNA (comprising of extracellular and fragmented DNA) would be diluted. The initial low concentration of PEG used in sample concentration and nucleic acid precipitation could possibly exclude low-molecular-weight DNA. In other words, the fragmented DNA could be lost in the initial concentration step.
Increasing concentrations of PEG and NaCl lead to the precipitation of lower-molecular-weight DNA
To test if increasing PEG concentration helps in enriching lower-molecular weight DNA, a laboratory standard of PCR fragments of different lengths was generated, creating a range of linear DNA fragments (Figure 4). The standard was subjected to four different precipitation conditions – 9% PEG-8000 + 0.3 M NaCl (the original combination), 9% PEG-8000 + 1.2 M NaCl (only raising salt), 30% PEG-8000 + 0.3 M NaCl (only raising PEG) and 30% PEG-8000 + 1.2 M NaCl (raising both salt and PEG). The conditions were chosen based on the standard protocol being used for SARS-CoV-2 surveillance17 and from conditions reported in a study of modular methods of DNA extraction from environmental samples33. Two different centrifugation speeds – 15,000 x g and 20,000 x g – were used based on the effects of differential speeds on DNA pelleting43. On increasing the PEG and NaCl concentrations to 30% and 1.2 M, respectively, the total recovery of DNA increased by approximately 70%, and DNA fragments as small as 150 base pairs (bp) were effectively precipitated (Figure 5). The difference in speed did not have much effect on the yield (Figure 6A) and may be due to the long centrifugation time used. Since the total DNA recovery was lower in the original PEG/NaCl combination, it was possible that lower molecular weight bands, although present, were at a concentration below the visualization limit. To test this, the amount of input DNA for precipitation was increased, and an excess of DNA corresponding to each treatment (1.5 mg) was loaded on the agarose gel for visualization. Only the treatment with 30% PEG showed good recovery of the lowest molecular weight, i.e., 150 bp band (Figure 6B).
Precipitation of circular DNA (bacterial genomic and plasmid DNA) does not follow the same trend
E. coli MG155 genomic DNA and plasmid (pRSV, ~4 kb) were used for precipitation with different PEG and NaCl concentrations. Unlike with linear DNA, there was no significant impact of raising PEG and NaCl levels (Figure 7), suggesting that yields of circular and high-molecular-weight DNA (typically intracellular DNA) are unaffected by raising PEG. This is in contrast to earlier observations with protein precipitation44, where solubility declined steeply with PEG concentration. Given that PEG acts as a crowding agent, it is hypothesized that the surface area of the macromolecule available for interaction plays an important role in its effectiveness as a precipitating agent. With linear DNA, as size increases, it is reasonable to hypothesize that the area available for interaction also increases. With circular DNA, it is possible that low PEG concentrations are already sufficient to saturate the molecule, and further raising the concentration may not increase the effective surface area for interaction.
Poor yield of DNA from wastewater when pre-processing precipitation with PEG and NaCl is omitted
To test if pre-processing wastewater is crucial for DNA extraction, 20 mL of heat-inactivated (70 °C, 4 h) wastewater was spiked with 10 mg of previously prepared linear standard and incubated overnight at 4 °C (a) without PEG + NaCl, (b) with 9% PEG + 0.3 M NaCl (original combination), and (c) with 20% PEG + 1.2 M NaCl (increased PEG and salt). The samples were then processed through the soil kit for DNA extraction, as explained in the protocol. It was found that the yield of extracted DNA increases by 60 % on pre-processing wastewater samples with 9% PEG + 0.3 M NaCl when compared to no pre-processing step (Figure 8), indicating that pre-processing PEG and NaCl precipitation is vital to obtain high-yield DNA from wastewater samples. It is also noteworthy that while the overall DNA yield, including high-molecular-weight genomic DNA (gDNA), is lower when high PEG and salt are used for precipitation, the proportion of lower-molecular-weight DNA is enriched (Figure 8). The decrease in overall yield on raising PEG and salt can be attributed to the highly viscous nature of PEG, which can lead to loss of DNA pellet while removing the supernatant. This strengthens the case for the proposed step-wise DNA extraction method (Figure 9), wherein high-molecular-weight DNA can first be extracted using low PEG and NaCl precipitation, and the resulting supernatant can be subjected to another round of pre-processing with increased PEG and NaCl to efficiently extract the low-molecular-weight DNA that escaped the first round of precipitation.
Figure 1: Pre-processing of wastewater samples. Schematic showing the workflow from sample collection to DNA extraction. Please click here to view a larger version of this figure.
Figure 2: Typical gel profile of DNA extracted from wastewater samples. A volume of 50-100 mL (as indicated in the figure) of wastewater sampled from different sites was heat-inactivated by incubation at 70 °C for 4 h. It was then incubated with 9% PEG and 0.3 M NaCl overnight at 4 °C and then processed to extract DNA using either soil or a water kit. Approximately 150 ng of total DNA extracted was loaded on a 1% agarose gel along with 1 kilo-base pairs (kb) ladder as a marker and subjected to electrophoresis (90 V, 30 min). DNA was visualized under ultraviolet (UV) light using the dye SYBR SAFE. The table on the right details sample collection sources, volume of wastewater processed, and kit used for DNA extraction for each lane. (STP: Sewage Treatment Plant; UVR: Ultraviolet Radiation) Please click here to view a larger version of this figure.
Figure 3: STP effluent samples show poor total DNA yields. A volume of 40 mL of wastewater sampled from STP influent and effluent was heat-inactivated by incubation at 70 °C for 4 h. It was then incubated with 9% PEG and 0.3 M NaCl overnight at 4 °C and then processed to extract DNA using either a soil or water kit or a bacterial genomic DNA extraction kit. Approximately 150 ng of total DNA extracted was loaded on a 1% agarose gel along with 1 kb ladder as a marker and subjected to electrophoresis (90 V, 30 min). DNA was visualized under ultraviolet (UV) light using the dye SYBR SAFE. The concentration of extracted DNA for the effluent samples was below 1 ng/mL of wastewater and hence could not be visualized. Please click here to view a larger version of this figure.
Figure 4: Generation of a linear DNA size standard to test the efficacy of low-molecular-weight DNA precipitation. Five different-sized DNA fragments were generated by PCR amplification using E. coli (MG1655) genomic DNA as a template and primers and conditions as described in Table 2. DNA purified (~1 µg ) with the PCR purification kit was loaded onto a 1% agarose gel along with 1 kb ladder as a marker and subjected to electrophoresis (90 V, 40 min). DNA was visualized under ultraviolet (UV) light using the dye SYBR SAFE. The lanes are labeled by gene names from which the fragment was amplified and the expected amplicon size. Please click here to view a larger version of this figure.
Figure 5: Low-molecular-weight DNA is efficiently recovered only at the highest (30%) PEG concentration. Input DNA (3.34 µg) from the size standard generated previously was treated with different combinations of PEG and NaCl as indicated in the figure and extracted as detailed in the protocol. Centrifugation speed used was 15,000 x g. For Input (lane 3) and 1.2 M NaCl + 30% PEG (lane 7), 0.8 µg of DNA was loaded onto the gel. For the rest, since the total yield was low, the total amount of DNA extracted in 16 µL was loaded and 1 kb ladder was loaded as a size marker onto a 1% agarose gel and subjected to electrophoresis (80 V, 45 min). DNA was visualized under ultraviolet (UV) light using the dye SYBR SAFE. The white box highlights the lowest band of 150 bp. Please click here to view a larger version of this figure.
Figure 6: Recovery of low-molecular-weight DNA is determined by high PEG rather than high salt concentrations. (A) Input DNA (14 µg) from the size standard was treated with different combinations of PEG and NaCl, as indicated in the figure, and extracted using ethanol precipitation as detailed in the protocol. Centrifugation speed used was 15,000 x g and 20,000 x g, as indicated in the figure. The entire amount of input DNA and extracted DNA, along with 1 kb ladder, was loaded onto a 1% agarose gel and subjected to electrophoresis (90 V, 30 min). DNA was visualized under ultraviolet (UV) light using the dye SYBR SAFE. (B) Input DNA (15.67 µg) from the size standard generated previously was treated with different combinations of PEG and NaCl, as indicated in the figure, and extracted as detailed in the protocol. The centrifugation speed used was 15,000 x g. Extracted DNA (1.5 µg) was loaded into each lane and 1 kb ladder was loaded as a size marker onto a 1% agarose gel and subjected to electrophoresis (80 V, 45 min). DNA was visualized under ultraviolet (UV) light using the dye SYBR SAFE. The white box highlights the lowest band of 150 bp. Please click here to view a larger version of this figure.
Figure 7: Recovery of genomic DNA and plasmid DNA is not significantly affected by increasing PEG and NaCl concentration. Input DNA (9 µg; 4.5 µg plasmid and 4.5 µg genomic DNA) was treated with different combinations of PEG and NaCl as indicated in the figure and extracted as detailed in the protocol. The centrifugation speed used was 15,000 x g. The entire amount of input DNA and extracted DNA, along with 1 kb ladder, was loaded onto a 1% agarose gel and subjected to electrophoresis (90 V, 45 min). DNA was visualized under ultraviolet (UV) light using the dye SYBR SAFE. The white boxes indicate the genomic DNA and plasmid DNA. Please click here to view a larger version of this figure.
Figure 8: Pre-processing precipitation with PEG and NaCl is crucial for high-yield extraction of DNA from wastewater. Heat-inactivated (70 °C, 4 h) wastewater (20 mL) was spiked with 10 mg of previously prepared linear standard and incubated overnight at 4 °C without PEG and NaCl or varying PEG and NaCl concentrations as indicated in the figure. It was then processed to extract DNA using a soil kit. Input DNA (4 µg) used for spiking (lane 1), and DNA extracted (4 µg) with pre-processing step of 9% PEG + 0.3 M NaCl (lane 3) was loaded onto a 1% agarose gel. For the rest of the conditions, the entire amount of extracted DNA was loaded on the gel since the yield was low. A 1 kb ladder was also loaded as a size marker. The gel was subjected to electrophoresis (70 V, 2 h). DNA was visualized under ultraviolet (UV) light using the dye SYBR SAFE. Please click here to view a larger version of this figure.
Figure 9: Proposed step-wise method of DNA extraction from wastewater to enrich both high and low-molecular-weight DNA. Flowchart depicting a two-step method of DNA extraction from wastewater with an initial pre-processing step using low PEG and NaCl concentration. The supernatant from the first step is subjected to another round of pre-processing precipitation with increased PEG and NaCl to effectively extract both high and low-molecular-weight DNA from wastewater. Please click here to view a larger version of this figure.
Table 1. Prior use of PEG and NaCl in studies for DNA extraction from environmental samples. Please click here to download this Table.
Table 2. Primer sequences and PCR conditions for standard generation. Please click here to download this Table.
Table 3. DNA yield and quality obtained from wastewater samples over a period of two months in 2023. Please click here to download this Table.
AMR is one of the top 10 health threats today, as listed by the WHO, and environmental surveillance for AMR is recognized as an important tool across the world. As mentioned in the introduction, a comprehensive record of environmental AMR includes low-molecular-weight, fragmented, and extracellular DNA. The pre-processing protocol reported here using a high concentration of PEG combined with salt (30% PEG and 1.2 M NaCl) achieves this result by enriching the proportion of low-molecular-weight DNA without impacting extraction of higher molecular weight fractions (largely genomic and plasmid DNA). This is in contrast to prior methods of concentration using lower PEG and direct kit protocols (without pre-processing) which are cumbersome due to the need for processing large volumes. Pre-processing with lower PEG concentrations was inefficient in recovering low-molecular-weight DNA ('low' refers to DNA bands under 600 bp here). When no PEG is used during pre-processing, DNA extraction is poor, resulting in only 10% of the input DNA being recovered as opposed to 70% when PEG and NaCl are used (Figure 8). In addition to the addition of PEG and NaCl, another important step is the removal of residual ethanol post-DNA extraction from the kit to get pure DNA for downstream processing. It is important to determine the initial volume of water needed to get a total amount of at least ~300 ng to 5 mg of DNA in a volume of 30-50 mL upon elution. In contrast with 1 L or more of water that is typically processed for DNA extraction from wastewater27,41, it was found that with the pre-processing step proposed here, a mere 27.5-40 mL suffices to get high-quality DNA for downstream processing. Wastewater also contains particulate and organic matter, which can contain adsorbed bacterial cells and extracellular DNA. Hence, the desorption of cells and DNA accompanied by cell lysis is an important step for increasing DNA yield. High PEG concentrations make the sample viscous and can hinder the lysis and desorption steps for DNA extraction, as mentioned previously. To avoid this, a step-wise method of DNA extraction has been proposed, as outlined in the flowchart (Figure 9). This method will be helpful in extracting both intracellular high-molecular-weight DNA (Figure 7), and extracellular low-molecular weight fragmented and circular DNA.
One current limitation of this protocol in terms of expense is the continued processing of enriched low-molecular-weight DNA through a kit for enhancing quality. This limitation can potentially be bypassed using other methods of DNA clean-up, like phenol-chloroform extraction. Another practical limitation is handling high PEG concentrations since the pellet can be lost due to the high viscosity of the solution. Currently, pre-processing using PEG and NaCl is carried out, followed by a modified version of an existing kit protocol, i.e., the kit is still used for final DNA purification. A kit-free purification of DNA, which gives a high DNA yield but of low purity (not reported here), is being optimized to improve the quality of the extracted DNA. Due to the viscous nature of PEG solutions, an optimal PEG concentration that can yield a full range of DNA sizes and yet be handled comfortably is to be aimed for. Lowering the concentration to 20% (Figure 6B) achieved a result intermediate to the extraction efficiencies of low (9%) and high (30%) PEG; this may be worth following up on in future work.
PEG is routinely used during different steps of DNA extraction from environmental samples. However, the use of PEG has not been standardized for DNA extraction in the context of AMR surveillance. Wastewater surveillance entails the detection of antimicrobial resistance in many different niches, including STPs (influent and effluent) and open and closed drains8,27. While the information on AMR in the influent provides crucial information on the resistance circulating in the community, the presence of AMR genes in the effluent is equally important to measure and potentially predict the emergence of resistance outbreaks. Effluent samples typically have low microbial load since most cells are killed during the treatment process, resulting in very low cellular DNA yield. However, effluents contain extracellular free-floating DNA comprising both high-molecular-weight genomic DNA and low-molecular-weight plasmid, phagemids, and fragmented DNA. AMR genes present in low-molecular-weight DNA (both fragmented and plasmid/phagemids) can get transferred to the remaining live cells in the effluent, leading to the dissemination of resistance30,41,45. Hence, it is important to detect low-molecular weight extracellular DNA in wastewater. Other methods employed for DNA extraction from wastewater while assessing AMR typically do not ensure initial enrichment of low-molecular-weight DNA. This approach of DNA extraction from wastewater can be used to provide information about a comprehensive resistome of the environment.
The authors have nothing to disclose.
We acknowledge funding support from the Rockefeller Foundation (Rockefeller Foundation Grant Number 2021 HTH 018) as part of the APSI India team (Alliance for Pathogen Surveillance Innovations https://data.ccmb.res.in/apsi/team/). We also acknowledge the financial aid provided by Axis Bank in supporting this research and the Trivedi School of Biosciences at Ashoka University for equipment and other support.
24-seat microcentrifuge | Eppendorf Centrifuge 5425 R | EP5406000046 | |
Absolute Ethanol (Emsure ACS, ISO, Reag. Ph Eur Ethanol absolute for analysis) | Supelco | 100983-0511 | |
Agarose | Invitrogen | 16500500 | |
Bench top refrigerated centrifuge | Eppendorf Centrifuge 5920 R | EP5948000131 | |
ChemiDoc Imaging System | BioRad | 12003153 | |
DNeasy PowerSoil Pro Kit | Qiagen | 47014 | |
DNeasy PowerWater Pro Kit | Qiagen | 14900-100-NF | |
dNTPs (dNTP Mix 10mM Each,0.2 mL, R0191) | Thermo Fisher | R0191 | |
DreamTaq DNA Polymerase, 5 U/µL + 10x DreamTaq Buffer* | Thermofscientific | EP0702 | |
E-Gel 1 Kb Plus Express DNA Ladder | Invitrogen | 10488091 | |
Maxiamp PCR tubes 0.2 mL | Tarsons | 510051 | |
Molecular Biology Grade Water for PCR | HiMedia | ML065-1.5ML | |
NanoDrop OneC Microvolume UV-Vis Spectrophotometer | Thermo Scientific | 13400519 | |
Parafilm | Bemis | S37440 | |
PEG-8000 | SRL | 54866 | |
QIAquick PCR & Gel Cleanup Kit | Qiagen | 28506 | |
Qubit 4 Fluorometer (with WiFi) | Thermofisher | Q33238 | |
Qubit Assay Tubes | Thermofisher | Q32856 | |
Qubitt reagent kit for ds DNA, broad range | Thermo Scientific | Q32853 (500 assays) | |
Sodium Chloride | HiMedia | TC046M-500G | |
SYBR Safe DNA Gel Stain | Invitrogen | S33102 | |
T100 Thermal Cycler | BioRad | 1861096 | |
Thermo Cycler (ProFlex 3 x 32-well PCR System) | Applied Biosystems | 4484073 | |
Wizard Genomic DNA Purification Kit | Promega | A1125 |